Here’s slides from my talk today.
For linear delta printers like the Kossel mini, the positional accuracy (assuming perfect mechanics and motors) isn’t the same uniform grid as on a cartesian printer.
The graph above shows the Z error when rounding the actuator (linear rail car) positions to the nearest 1mm on a printer the size/dimensions of a Kossel mini. Near the center there’s less error (more white), and near the towers you can see strong red/blue rings where that tower’s rounding is more prominent in the position of the effector (the part your hotend is mounted to).
In the same way that whosawhatsits microstepping calibration test shows rounding on a cartesian printer, I’m sure there’s a mechanical way to see these rounding problems on a print. Maybe just a 6 point asterisk pointing towards and away from towers would show them.
Just thought that was a neat thing to graph.
I’ve rebuilt one of my Prusa i3 printers into a RepRap Wilson:
It’s… fine. The Wilson has more build height and looks cooler, and has a better Y axis (except giant printed parts). Unfortunately it isn’t what I was looking for to have an easier printer to do a build weekend for. It’s not easier to assemble than an i3, and printing the parts is about the same or more time consuming.
The hackerspace now has access to a laser cutter that can do 6mm melamine without problems so I want the next printer we have a build weekend for to have more laser cut parts. I’d also like to build a Core XY machine. Here’s what that looks like so far.
For my 4th 3d printer I wanted the speed of a delta, and dual extruders for soluble support material or multi color or whatever. (For those keeping track at home the order has been MakerFarm Prusa i3, Mendel 90, Prusa i3 edge, Kossel mini dual-extruder, not discussed yet).
This has been quite a journey. I started on this in [mumble] and am still not quite there yet.
Ok, so getting Kossel mini vitamins and printing the parts is easy. (Misumi takes forever to ship 1515 from overseas but whatever)
Building two e3d v5 1.75mm bowden hotends and wiring them up to RAMPS is easy.
Configuring marlin for a delta and dual extruders is easy.
Redesigning the Kossel mini end effector to hold two e3d hotends was… possible.
And it turns out my dual spool holder idea works pretty well.
But printing with dual extruders on a delta found a bug in Marlin where it would try to move in cartesian space (not delta space) after changing tools, which would attempt to destroy the machine. And sort of did.
[The very simple looking belt holders on the standard carriage shear off rather than letting the belt slip]
I fixed that in Marlin but never quite got the nozzles aligned and everything calibrated right.
Since then I’ve moved from airtripper extruders to RichRap’s greg’s wades variant, and to a 5 axis Smoothie board. I’ve also replaced my open loop belts with closed loop ones as in the standard BOM. (I didn’t find a closed loop variant of the carriages that worked for me)
Here’s my bulleted list of current problems
- Solvable: The electronics don’t have an enclosure now so it’s not really a printer that can be moved.
- Solvable: I haven’t ever finished calibrating the position of the two nozzles because I end up changing how the hotends are mounted before I get that far.
- Hard: The nozzles are very difficult to align to the same height, so typically one will mess up the printing of the other.
- Hard: With the extra weight of the two e3d (v5) hotends and fan duct, even an acceleration of 3000 causes very visible ringing at corners. I THOUGHT THIS WAS SUPPOSED TO LET ME GO FASTER!
Sorry I don’t have more good pictures.
The 12V embedded supply that has been doing very nicely on my Mendel 90 ate itself rather with a very enthusiastic pop and the usual smell of burnt electronics.
Thankfully the fuse inside the supply caused it to fail safely (power shut off to the supply input and nothing else was going to burn up), but it means I needed to quickly find a new power supply.
More knowledgeable people than myself suggested it was probably something else that failed. There was a weird mark on the rectifier (and my thinkpad brick died at the rectifier)
But I think this was the source of the major failure. One of the output mosfets failed blowing black burn marks out towards the small transformer.
I have a couple spare 12V server power supplies lying around, and it turns out they convert easily for custom use. See this thread: http://www.rcgroups.com/forums/showthread.php?t=1292514
My new supply was a Dell AA23300, which there are pinouts for turning on here: http://www.tjinguytech.com/my-projects/server-p
So, with a cartesian X/Y gantry backlash (as caused by a loose belt) on one axis results in circles with a flat on one side. Here I’m showing the desired motion (red circle) and actual motion (blue) of a 100mm circle with 5mm of backlash in the X axis.
The flat sides happen on the parts of the circle where the X axis motion is entirely lost due to the backlash.
With backlash on both the X and Y belts, you get 4 flat sides instead of 2.
OK, so what happens on a delta, like the Rostock? Here’s 5mm of backlash on the same circle, but applied to the delta tower centered above the image (The 3rd tower in the Kossel solution). The resulting “Flat” isn’t flat because the motion of the one axis doesn’t have a direct impact on the movement of the print head.
With backlash on all 3 towers we see a hexagon of flat spots.
Just like with the cartesian (not shown), the corners move around a little when the direction of the circle changes:
I hadn’t considered before that the print direction changing will show defects at different places due to backlash. This suggests a slicer could be modified to print perimeters in alternating directions to emphasize backlash issues.
After a few months of frustration with the Magma hotend, which can print ABS and nylon but not PLA, I decided to try the E3D hotend.
Unfortunately, it didn’t just drop in. The heatsink on the E3D hotend doesn’t have the same groove as the Magma (the magma groove matches a J-Head). The lip on the top is thinner and the groove is larger. With the Mendel 90 wades extruder, the groove thickness isn’t important but the lip being thinner means that the E3D hot end isn’t fully secured with the narrower lip. I modified and printed a new wades extruder by changing to “function jhead_groove_offset() = 3.8;” in scad/vitamins/hot_ends.scad
The included 30mm fan duct also doesn’t fit on the Mendel 90 right next to the wingnuts. I redesigned the X carriage in the same vein as this modified one on thingiverse. My X-Carriage uses LM10UU so I couldn’t go with that x-carriage.
I created a new fan duct in OpenSCAD inspired by the used with the X-carriage above, except mine ducts the fan directly to the E3D fan heatsink. It took a few iterations to get the shape correct and fitting, and with reasonable support material. The Cura auto-generated support material was fine for the print, but I broke the duct in half trying to remove it.
My first print was ABS because it’s what I had on the printer. PLA also prints well. I figured I’d test my 2nd PLA print with this hotend by printing at 60 microns (0.06mm layer height).
This was printed at 50mm/sec using a fan off to the side of the printer. I need to get a fan duct if I’m going to do PLA much more.
So far this hot end has been working well for me. I perhaps have too much cooling on it with the high-airflow 40mm fan. The increased length has reduced my Z height by about 12mm though.
My Mendel90 modifications for E3D hotend and LM10UU are now on thingiverse.